It was a bit of a struggle to record the motion capture and playing back live the body animation, so that body and hands match time wise
I was anticipating this, but haven't thought of a solution yet. If you eventually come up with ideas, let me know.
Maybe constraint hand and a camera to the mocap wrist, so you can see it in place?
---
Speaking of body, I've also touched MediaPipe Body in VML Streamer, to make it support world transforms as well. (previously it only had hips fixed at the origin). I can now record jumps (although with some Z jitter). Also added support for more than 1 body. (I've been able to capture up to 4-5 bodies, as soon as they don't overlap and camera is fixed/has decent lighting).